VOL.29
NOTES
2440
proved to be identical with cyclohexyloxindole from the catalytic reduction (part A) by mixture melting point and infrared spectral comparison. Reduction of 1soindigo.-A hot suspension of isoindigo (0.5 g.) in 2-propanol (40 ml.) and water (5 ml.) was treated with sodium borohydride (0.12 9.). Within 10 min. all of the solid had dissolved, and the boiling solution was colorless. More water (20 ml.) was added, and the mixture was kept hot 20 min. longer. The solution was concentrated under reduced pressure to onethird of the original volume, and a solid began to separate. The suspension was reheated and the solid was dissolved by addition of alcohol. On standing, the hot filtered solution deposited colorless crystals, 0.5 g., m.p. 258". Recrystallization from alcohol raised the melting point to 272-274'. Mixture melting point and infrared spectra showed the product was identical with a sample of leucoisoindigoprepared by zinc-acetic acid reduction of isoindigo.l*
Long-Range Deshielding by Chlorine RAYMOND C. FORT,JR.,'GORDON W. H. CHEESEMAN, AND EDWARD C. TAYLOR
Frick Chemical Laboratory, Princeton University, Princeton, New Jersey Received December 6, 1963
Long-range shielding and deshielding effects have been known to n.m.r. spectroscopists for some tinie.2 Most of the clearly defined examples involve the anisotropy of unsaturated functions. We report herein the observation of an unusually large chemical shift produced by a chlorine atom five bonds distant from the shifted proton. Compounds 2-5, Table I, having R8 = C1, all show an absorption in the range 0.97-1.20, integrated intensity of one proton. This peak has been assigned to H, on the aromatic ring. It appears as the X portion of an ABX system: an intense quartet and two weak combination bands.34 The AB portion of the spectrum, H2 and Ha, is coiiiplicated by overlap and/or coupling to HI, so that the values for J A X and J B X cannot be obtained. However, the sum ~ J A X J B X l = 10.0 C.P.S. can be measured. Upon reductive removal of the chlorine atom (conipound 1) the Hq peak iiierges with the Hz H, resonance a t ca. r 2.0. A long-range deshielding effect, produced by the anisotropy of the carbon-oxygen double bond, previously has been noted for the carbonyl group in formally similar structures. For example, Crombie and Lown5 have used thc shielding of the 1-hydrogen by the 12-carbonyl to assign the geometry of the B/C ring fusion in a series of rotenoids (I). An& has found that restricted rotation of the N-acetyl group in strychnospermine (11) produces a downfield shift of
+
I
n
the aromatic proton a t C-12. At room temperature, rotation of the acetyl group is sufficiently rapid that the two possible environments for C-12 are averaged out. At lower temperatures, that conformation having the carbonyl adjacent in space to C-12 displays the unusual shift. The geometry of this frozen conformation closely resembles that of compounds 1-5. A few reports of such deshielding by chlorine are scattered through the literature. Huitric' recently has suggested that, in cis-2-(o-chlorophenyl)cyclohexanols (111), the o-chlorine causes a downfield shift of the 1-hydrogen resonance. However, the shift is snialler by an order of magnitude than that observed in this work (0.1 p . p m us. 0.8-1.0 p.p.ni.). The spectrum of $7-dichlor~quinoline~ (IV) shows a downfield shift of 0.48 p.p.m. for Hd when compared to 5,7-dimethylquinoline. In this instance, the chlorine and the shifted proton are four bonds apart in a conjugated system; a part of the effect may be transmitted through the ?r-system. Similarly, the indicated aromatic protons of Vg show a downfield shift of 0.52 p.p.m. when compared to the aromatic protons of 9,lOdihydroanthracene. Finally, Osawa and Neeman'O report chemical shifts for H1 in the estrones (VIa and b) of r 2.82 and 3.14, respectively, a difference of 0.32 p.p.m.
@&7
cr
OH
I11
H 4
&l
c1
IV
+
(1) Allied Chemical Corp. Fellow, 1962-1963; National Science Foundation Predoctoral Fellow, 1963-1964. (2) I,. M . Jacknian. "Applications of Nuclear Magnetic Resonance Spectroscopy in Oreanic Chemistry," Pergamon Press, New York, N. Y., 19,59, p. 112 ff. (3) .J. .I.Pople, U , G. Schneider, and H. J. Bernstein, "High Resolution Nuclear Xfaanetic Resonance," hIcGraw-Hill Book Co., Inc., New York, N. Y . . 1959, p. 137. ( 4 ) Strictly speaking, Hd is the X part of a n .II3CX syatem: however, sinre H I does not appear to he coupled t o Hi. ARX is a reasonable description. (5) I,. Cromhie and J . W. L o n n , Proc. Chem. Sac. (London), 299 (1961). (6) F. A . L. .Inet, Can. J . Chem.. 4 1 , 883 (1963).
c1 V
It is possible that the great magnitude of the shift observed in our pyrrolo [1,2-a]quinoxalines results from both electrostatic electron withdrawal (a field effect) and the anisotropy" of the carbon-chlorine bond,12 rather than the latter alone. Either effect should be enhanced by the close approach of the two atoms; (7) A . C. Huitric, J . Ore. Chem., 97, 715 (1962). ( 8 ) F. A . L. Anet, J . Chem. Phy8.. 89, 1274 (1960). (9) Varian Associates. "High Resolution N M R Spectra Catalog," Spectrum No. 338. (10) Y .Osawa and M. Neeman, J . A m . Chem. Soc., 86, 2856 (1963). (11) Ref. 2 , p. 115. (12) G. 9. Reddy and J. H . Goldstein, J . Chem. Phye., 38, 2736 (1963).
244 1
NOTES
AUGUST, 1964
TABLE I" N.M.R.SPECTRA O F
Compound
RI
R9
R3
Ri
PYRROLO [ 1,2-U]QUINOXALINES
Rn
R3
r-values HI
Ha
-
H3
H4
HI
CBH~ CHI H -2.5 7.74 2.6-2.8 2.6-2.8 2.02 4 3.29 2 CH3 H c1 7.42 3 30 ... 2.27 - 2 . 7 1 -.+ 1.16 3.43 3 CHJ CH 3 c1 7.48 7.91 ... 2.30 2.75 -t 1.20 3.48 4 CaHa H c1 2.60 3.17 ... +2.13+ 1.02 3.38 5 C6H5 CHI c1 2.54 7.73 ... 2 . 0 6 -.+ 0.97 3.19 The preparation of these pyrrolo[l,2-a]quinoxalinesis described elsewhere [E. C. Taylor and G. W. H. Cheeseman, J . A m . C h m . Soc., 86, 1830 (1964)l; a preliminary account of the synthetic work was presented a t the 145th National Meeting of the American Chemical Society, New York, N . Y., Sept., 1963, Abstracts, p. 70Q. 1
+
+
measurements on models give a C1-H distance of only 2.27 A. The variation of the position of the H1 resonance is also worthy of attention. In compounds 2 and 3, where R1 = methyl, HI appears a t lower field than Hz and Ha, reflecting its position ortho to nitrogen. I n compounds 4 and 5, however, when R1 = phenyl, H1 is shifted to higher field, merging with the H2-H3 resonance. We attribute this shift to conjugation of the 1-position with the phenyl group; resonance structures can be written which place a formal negative charge a t position 1. All spectra were obtained on a Varian Associates A-60 spectrometer a t normal operating temperature. Compounds were run as ca. 10% solutions in reagent grade carbon tetrachloride with 1% tetramethylsilane as internal standard. Chemical shifts are considered accurate to j=l c.P.s., coupling constants to *O.l C.P.S. Acknowledgment.-We wish to thank Dr. Pierre Laszlo for a helpful discussion. This work was supported in part by research grants to Princeton University from the National Cancer Institute, National Institutes of Health, U. s. Public Health Service (Grant No. CA-02551), and from the American Cancer Society.
Base-Induced Fragmentation of 2-Phenyl- 1,3 -dioxolanel P. S. WHARTON, G. A. HIEGEL,AND S. RAMASWAMI Department of Chemistry, University of Wisconsin, Madison, Wisconsin Received February 13, 1964
We wish to report a representative of an unexplored and potentially useful class of reactions : 1,a-eliminations of anionic fragments.2-4 (1) Grateful acknowledgment is made of financial support from the Petroleum Research Fund (Grant 1116-A4). (2) An interesting and related example was reported by R. L. Letsinger and D. F. Pollart, [ J . A m . Chem. SOC.,18, 6079 (lQ56)l. They found t h a t treatment of 2-phenyltetrahydrofuran with propyl- or phenyllithium, followed by work-up, yielded ethylene and acetophenone in high yields. Unpublished work with Y. C. Poon has furnished another variation: ethylene glycol sulfite is reduced to ethylene in variable yield (BO far not greater than 30%) by metallic sodium or potassium in refluxing xylene.
2-Phenyl-l,3-dioxolane1treated dropwise,with phenyllithium in ether a t room temperature, rapidly evolved a gas which was collected (80-94% in three separate runs) and shown to be pure ethylene by mass spectrometry. Formally, the reaction proceeds by 1,3elimination of benzoate anion from the conjugate base of 2-pheny1-lJ3-dioxolane. Although benzoic acid was not recovered from the reaction mixture the observed products are consistent with the elimination of lithium benzoate and subsequent reaction with phenyllithium : from one run benzophenone and triphenylcarbinol were recovered in an over-all yield of 85% based on 2-phenyllJ3-dioxolane. Experimental The reaction vessel, a 250-ml. three-necked round-bottomed flask, was fitted with a gas inlet tube, addition funnel with pressure-equalizing side arm, and a condenser. The condenser was connected to a gas buret via two traps cooled in a Dry Iceacetone mixture and a gas sampling tube. The Dry Ice traps prevented ether vapors from reaching the gas buret and thus facilitated the volumetric and mass spectroscopic determinations of gas evolved in the reaction. The entire system was thoroughly flushed with helium, both before and after adding 4.892 g. (32.6 mmoles) of 2-phenyl-1 ,3-dioxolane6 to the reaction vessel and 100 ml. of an ethereal solution of phenyllithium (ca. 125 mmoles) to the addition funnel. After allowing the volume of the system to equilibrate (ca. 1 hr.) the solution of phenyllithium was added dropwise to the reaction vessel until gaa evolution ceased. The reaction mixture was stirred (magnetically) continuously and the reaction vessel was immersed in a bath of water a t room temperature. Gas evolution ceased when 17 ml. of phenyllithium solution remained madded. The increase in volume, 769 ml. a t 736 mm. and 22O, corresponded to 94y0 of the theoretical evolution of 1 equiv. of gas. The mass spectrum of the gas ( m / e = 70 to 12) was indistinguishable (*2'%) from a maas spectrum of pure ethylene obtained under the same conditions. (3) 1,3-Eliminations of stable and unstable neulrol molecules are well known and are referred to in the exhaustive documentation of 1.3-additions by R. Huisgen, Angew. Chem. Intern. Ed. Engl.. 9, 565 (1963). (4) The conversion of 1,2-diols to olefins provides an example of immediate synthetic applicability. C/. E. J. Corey and R. A. E. Winter, J . A m . 86, 2677 (1963). Chem. SOC., (5) H. Hibbert and J. A. Timm. ibid., 46, 1283 (1924).
